Non-Factorised Variational Inference in Dynamical Systems

Abstract

We focus on variational inference in dynamical systems where the discrete time transition function (or evolution rule) is modelled by a Gaussian process. The dominant approach so far has been to use a factorised posterior distribution, decoupling the transition function from the system states. This is not exact in general and can lead to an overconfident posterior over the transition function as well as an overestimation of the intrinsic stochasticity of the system (process noise). We propose a new method that addresses these issues and incurs no additional computational costs.

6 pages, 1 figure, 1 table

References in corpus (2)

• Variational Gaussian Process State-Space Models
• Closed-form Inference and Prediction in Gaussian Process State-Space Models

Cited by in corpus (5)

• FedLoc: Federated Learning Framework for Data-Driven Cooperative Localization and Location Data Processing
• Structured Variational Inference in Unstable Gaussian Process State Space Models
• Learning Nonlinear State Space Models with Hamiltonian Sequential Monte Carlo Sampler
• Bayesian Hidden Physics Models: Uncertainty Quantification for Discovery of Nonlinear Partial Differential Operators from Data
• Learning While Navigating: A Practical System Based on Variational Gaussian Process State-Space Model and Smartphone Sensory Data